Accelerate Your Product Management Workflow with AI

Welcome, Introductions, and Poll #1 (Your AI Experience)

Cameron Lanier | 00:00:00–00:04:00
my name is Cameron Lanier. I am the director of content at Productside and today you won’t hear too much from me, which is probably good news for you.

you’re going to be hearing quite a bit, however, from Dean Peters, who many of you already know. he is one of our principal consultants and trainers here at the Productside. he’s also kind of our resident AI expert.

here at Productside, if you don’t know too much about us, we focus on transforming your product management teams through training, through our courses, and we focus heavily on outcome-focused product management training, which you’ll hear a little bit more about today.

joining us today from Stormboard is—let me get this right—Reg Cheramy.

Reg Cheramy | 00:04:00–00:04:15
nailed it, nailed it.

Cameron Lanier | 00:04:15–00:05:30
okay, so Reg is joining us. he’s the founder and CEO of Stormboard, which, if you’re not familiar with it, is a digital whiteboard platform. there are some very exciting features they’ve recently added. it’s a great collaborative tool for product managers to use.

so, with that said, let’s talk about a little bit of housekeeping real quick.

one of the main questions we’re going to get asked is: will this be made available to watch later? yes. you will get an email, if you have already registered, with the information on how you can watch this later and watch it on demand at your pleasure.

beyond that, we also have the chat going, as you can already see. that chat will be monitored and we will be taking questions throughout via the Q&A—that is our preferred way to do that—so please feel free to send us questions. we’ll take those, and we’re going to try to reserve some time at the end for questions as well.

with all that said, we want to start things off by getting to know a little bit about you. so we want to start with a poll question.

Cameron Lanier | 00:05:30–00:07:30
our first poll question is: how many hours, how much time have you spent in ChatGPT or with similar tools?

so yes, we’re already getting answers really quickly here.

if you’ve spent:

no time

1 to 10 hours

10 to 50

50 to 200

or you might be like Reg and Dean here, who’ve spent 200+ hours

I’m not sure they want to admit it, but we all know it’s true.

we’ve got about 72 percent of you already participating, so thank you for that.

yes, in total—how many hours have you spent in ChatGPT or similar AI tools, whether it be Bard or what have you.

all right, we’re slowing down. I’m going to count down: five, four, three, two, and one. we’re ending the poll.

I’m gonna share the results.

so let’s take a look. I’m surprised, guys, looking at this, how many people filled in “none,” with us having an AI-centric webinar here. we have quite a few people who have not spent time with it at all.

so I guess we have people who are here to learn. anything surprise you or interest you in these results?

Dean Peters | 00:07:30–00:09:00
I am, a bit—considering all the chatter I hear and of course all the, what I call, prompt-engineering parlor tricks that I see.

you know, the “top 10 prompts for product managers,” and of course they have no context, so people use them and then go “meh,” as opposed to those of us spending more time with it.

I don’t know, Reg, what’s been your experience with that?

Reg Cheramy | 00:09:00–00:10:00
yeah, I think the interesting thing about the numbers here is they’re pretty evenly split between “none,” “1 to 10,” and “10 to 50.” it’s really neck and neck between those three.

so we’ve got a nice broad array of people that have used this.

and then, as expected, the crazy ones like you and I in that “50 to 200+,” that’s where it starts to drop off.

it’s really interesting to see that there’s a lot of people who have put 10 to 50 hours into trying to figure out how to make ChatGPT work for them.

I’m interested in the results of the next poll that we do near the end, that helps us understand a little bit more as well.

Dean Peters | 00:10:00–00:11:30
yeah, and I like your point there—it looks pretty even.

so let’s tell a story, right? if you know anything about any of us at Productside, we like to teach through story.

we like to talk about not the smoke jumper’s paradise, but the smoke jumper’s paradigm.

so Reg, you’re in control of the board, and you notice that I’ve used a little bit of prompt engineering.

the first one is: how many times—put a number in the chat—have you, as a product manager or product owner, been parachuted into a new situation?

whether it’s:

a new product manager job,

parachuted into a new fire,

parachuted into a new product,

I’m seeing “too many to count” start to come in.

we all understand this metaphor—being parachuted in or dropped in.

and so I decided to show some ways that we get parachuted in.

for example, imagine you’re that smoke-jumper-like product manager. you’ve taken training from us at Productside, we’ve given you all sorts of tools.

you’re like that new smoke jumper: “hey, I got a neat new axe, a neat new gas mask, and hey, look at this, we’re in a plane, isn’t this cool… hey, why is the door in the plane opening? oh, I get pushed out.”

now you’re falling to earth with the parachute unfurled, and you’re looking down at the ground saying, “I don’t want to land in the lake, those trees look awfully pointy, and I definitely don’t want to land in that fire.”

how do you land?

as a smoke jumper, first thing you have to do is land on target, check your equipment, check yourself. you’ve got to talk and collaborate, have a shelter ready, have radio contact with your partners, and then make sure you’re using the right tool for the job.

so if you look at the other set of actions—these 12 actions—these are all tools that you’re going to need to know how to use and when.

we thought it might be good to take a smoke-jump situation around: you’re at a job, it’s been a three-day weekend, or you come back from vacation and you see a press release from your CEO.

the press release says they want to create an AI-driven pizza delivery ordering service, and this is the first time you’re reading it—you’re parachuted in.

and by the way, a Slack message comes up saying, “hey, did you see the press release? you and your team get to go ahead and write this app,” because the CEO was tired of taking her family on vacation and not being able to figure out how to get gluten-free pizza in a town she’s never been to before.

and that takes us to the first section in our board.

The Smoke Jumper Metaphor and Jobs-To-Be-Done Setup

Dean Peters | 00:11:30–00:15:30
this means your CEO—just like your stakeholders, just like your customers, just like your engineers—comes to you in solution-speak.

she’s saying “build me an app that does this.”

all of these people come to us in solution-speak.

it’s your job, as a product manager or product owner, to get that conversation back into the problem space.

and the first place to do that is in jobs to be done:

what is the end user trying to do?

what is the end user feeling in terms of pains?

what do they want in terms of gains?

when I did this with some prompt engineering in ChatGPT, it took me about 500–600 lines of prompting.

but now, Reg, why don’t you go ahead and pick up this press release and—without all those lines of prompt engineering—show us how we can start scaffolding a conversation around outcomes.

if you’ll notice what Reg has here: he has these templates in these sections. those are the templates we teach with outcome-focused product management at the Productside.

he’s taken these templates, put them into Stormboard, and he’s going to show you the next generation of AI tools to help you get more product management done instead of more prompt engineering.

Reg, take it away.

Stormboard + Storm AI Demo: Jobs, Pains, and Gains

Reg Cheramy | 00:15:30–00:22:30
awesome, thanks Dean.

here, we’ve got a jobs-to-be-done template. as Dean mentioned, we’ve mirrored that inside Stormboard.

I’m just going to drill into it. what you’re going to see are the three sections that were referenced there:

customer jobs

pains

gains

when you’re doing a jobs-to-be-done exercise, there’s oftentimes initially a struggle: how do I get started, how do I prime the pump, how do I figure out what is a “job?” what’s a pain, what’s a gain?

what we’ve done with Stormboard is we’ve created a product edition called Storm AI that brings AI into the conversation.

what we’re doing is super unique because it understands the context of the “storm” (the board).

I’m going to go into settings here, into Storm AI, and we’re going to see this is that press release—the summary from the press release—that you guys experienced in the April 12th webinar.

we took and just copied a couple of the paragraphs, just to give Storm AI a little bit of context.

Dean Peters | 00:22:30–00:23:00
yeah, not even the whole thing, it’s not even the whole thing.

Reg Cheramy | 00:23:00–00:25:00
exactly. and it’s not a prompt; it’s just literally some text that describes the situation.

now, when we go out of this and click the Storm AI button, what it’s going to do is use that context and automatically give you customer jobs.

that’s the only context—just to be clear. the only context we gave it was literally that two-paragraph summary.

then, because Storm AI understands the template and the structure here, it’s able to give us some customer jobs that are perfectly in line.

now I know, Dean, when you were playing with this for your past webinar, you spent hours behind the scenes getting this all ready to present.

yet with Stormboard, you’re able to literally click a button and get customer jobs—no prompt engineering required. you just hit that little light bulb and ideas come your way.

Dean Peters | 00:25:00–00:26:00
yeah, and think about this when we show you these templates.

at Productside, we teach these as conversation centerpieces.

we want to use generative AI to start great conversations and to scaffold great sessions—to steal a term from Ruby on Rails.

it’s generated ideas, and if you notice, I dropped in another job. which means, I think, Reg, you could talk to that: now that we’ve jump-started this conversation, think about how much easier brainstorming will be if you have this type of contextual conversation starter with the customer jobs.

and I think there are a couple of others here that we wanted as well.

Reg Cheramy | 00:26:00–00:27:30
yeah, exactly. so I’m going to click on “pains” here.

what’s important to note is: this is a collaboration between AI and your teammates.

this isn’t about just having AI give you all the answers—although it does a pretty damn good job of giving great ideas.

this is about having a collaboration, because the more ideas you add or delete—if some of these ideas didn’t make sense, you can delete them—then Storm AI is going to understand all the ideas in there, both human and AI-generated, and give better ideas in each section as it goes.

it’s really a virtuous cycle, where we’re using the collaboration powers of Stormboard to train the AI in real time based on the ideas in each section of the storm.

we’ll fire up some gains as well.

the great part here is we’re able to just get some ideas.

one of the debates that we often have over at Stormboard is: should we start just with the team’s ideas before we add AI, or should we prime the pump with AI?

I think you’re going to experiment with that as a team because there are different outcomes from going both ways.

even if you start with the team, then you layer on the AI, the AI is going to give you even better ideas, and then you can bring humans back and have this back and forth.

Dean Peters | 00:27:30–00:30:00
and I’ll add to that “where do we start?”

for example, if you’ve listened to Teresa Torres’ book on continuous discovery, she talks about brainstorming and mentions that studies have shown people sometimes socially loaf or dominate the conversation at first, injecting a lot of bias.

you can avoid that injection of bias by having these conversation starters here.

it looks like we’ve got some good pains and gains.

maybe it’s a good time to see if we have some Q&A questions.

Cameron Lanier | 00:30:00–00:31:00
yeah, so we have one question. this is for Reg.

somebody asked about a waitlist: when will Storm AI be available?

Reg Cheramy | 00:31:00–00:32:00
yeah, so we’ll talk a little bit about how you can get access to Storm AI at the end of this call.

anyone that wants to give it a try—we have the ability to try it now, and it’s available for our enterprise customers immediately.

we’ll talk a little bit more about that at the end of the call.

Problem Space, Prioritization, and Data Security Questions

Dean Peters | 00:32:00–00:34:00
hey Reg, want to do a PowerPoint presentation on this?

Reg Cheramy | 00:32:00–00:34:00
awesome.

one of the beautiful things about Stormboard is our templates are dynamic. instead of just being a picture in the background, all these sections are dynamic, so we can do great things like reporting.

you can now generate a PowerPoint report of these jobs-to-be-done, because you don’t want to take a screenshot and send it to your stakeholders.

Dean Peters | 00:34:00–00:35:00
I’ll tell you what I have to do right now—even with plugins I have to use all sorts of markdown voodoo magic in ChatGPT or Bard to get it in a format I can import into PowerPoint, and I still have to massage it afterwards.

so now you get not only great conversation pieces but communication pieces.

this raises the question: do you want to spend more time prompt engineering or doing product management—talking about the problems we’re solving?

now that we’ve got some pains, gains, and jobs-to-be-done, I think we should keep moving this conversation forward for our friends who have been smoke-jumped or dropped in the middle of a fire of a new initiative.

Reg Cheramy | 00:35:00–00:38:00
awesome. we’re going to go back up to the main webinar storm and talk next about the problem space.

we’ve kind of got our jobs-to-be-done; now, Dean, talk a little bit about the problem space.

Dean Peters | 00:38:00–00:40:00
you’re actually seeing two templates here.

one is, again, a brainstorming exercise on the problem space: how do we get people thinking about the problems before we start talking about the solutions?

I think, Reg, you and I were joking the other day: three or four years ago the hippo comes running into the room saying “give me web 3.0,” then “give me blockchain,” then the seagull manager swoops in and says “give me NFTs,” and this past year you have the rhino charging into your office saying “give me generative AI.”

we always get this solution-speak.

what I like about the next-generation tool like Stormboard is it can reverse-engineer that solution-speak into problem space first through brainstorming, and then use a very simple prioritization process where we’re basically just saying:

is it high value and low cost (what we want)?

is it low value and high cost (what we want to avoid)?

or is it high value, high cost or low value, low cost where we need nuance—split the work, reconsider, etc.?

so, Reg, show us how to reverse-engineer that solution-speak into a conversation piece around the problem space.

Reg Cheramy | 00:40:00–00:44:30
awesome. once again, what I’ve done is combine those two templates into one, just to make it an easy conversation.

I’m going into the Storm AI context, and you’re going to see I’ve taken a high-level pitch of this. instead of the multi-paragraph context we used in the last one, this is something really simple.

again, you’re not entering a prompt; you’re just describing the scenario.

Storm AI will generate problems in the problem space automatically:

limited delivery options,

unclear ingredients in the pizza,

difficulty communicating dietary restrictions…

all of these are key problems related to getting gluten-free pizza in a town that you’ve never been to.

Storm AI is just instantly, almost magically, generating problems.

obviously, you’re going to want your team to add their own problems too, inspired by this, creating ideas.

then, once you’ve got a bunch of ideas, you want to prioritize them, because not all problems are equal.

some are high value, some are low value; some are low cost and low risk, some are high cost and high risk.

we can drag something like “fear of cross-contamination” into the high-value portion; maybe it’s high value and low cost to address.

and because it’s collaborative, if I don’t agree with where someone put an item, I can drag it somewhere else.

the struggle you might have is that ideas are overlapping because there are so many in that high-value area, but the template is dynamic—we can resize it and Stormboard keeps all the ideas proportionately placed.

Dean Peters | 00:44:30–00:46:00
normally, when I’ve seen people doing this without the benefit of generative AI, everything tends to go high and to the right, which means “do everything, everything’s priority one.”

what you want to do is look within that list and find the thing that goes dead center, use that as your point of reference, then everything becomes a relative prioritization.

the same thing would happen in the high-value/high-cost quadrant: if something is high value but high cost, maybe think about splitting it. it might reduce the value a bit, but if you split it, you can get some pieces closer to that quadrant one.

Reg Cheramy | 00:46:00–00:48:30
again, the great thing about having this template is: when you look at all these high-value, low-cost problems, it’s going to inspire you to think of more problems that are similar.

we might look at some problems and say, “these two are location-related,” and group them.

that’s that process Dean mentioned—affinity mapping.

the same pattern shows up if you look at the upper-left quadrants; all of this is about an iterative collaborative process.

then Storm AI is just that generative piece that helps you think of problems you might not have thought of.

one of our customers said one of the things they love is that sometimes Storm AI gives weird ideas—“bad” ideas. that’s actually great, because it encourages your team to think more wildly: “my idea can’t be worse than the AI’s.”

it frees up people to think more creatively.

we can also export this as an Excel report.

because now we’ve got these prioritized problems, we might want to put them into another tool, and oftentimes Excel is a great place.

you’ll see all the ideas, including unique identifiers. it knows the color, shape, x-y position, what section, and which quadrant each idea is in.

so it knows this one is in the high-value, low-cost quadrant, and it will rank it from 0 to 10 on each axis.

you can take that data—actual quantifiable data—and slice and dice it to make even better decisions.

Solution Space: “How Might We” and Poll #2 on AI Effectiveness

Cameron Lanier | 00:48:30–00:49:30
yeah, we got a couple questions.

first, I’m glad you showed the Excel, because that was one of the questions.

we also have a couple of concerns about confidentiality. some people asked about data security: is data entered in the system available to everybody? someone gave a warning about feeding models with confidential information.

can you speak to that a bit, Reg?

Reg Cheramy | 00:49:30–00:52:00
100%. our customers are large enterprises primarily, and their data security and privacy are paramount.

when you use ChatGPT on the web or in their app, they’re using your data to train their models.

with our solution, they don’t—because it’s using the API, and we’ve got a dedicated instance on the Azure OpenAI service.

the only thing that happens with your prompts is they’re logged for security auditing and compliance, and absolutely scrubbed in a maximum of 30 days.

they’re not used for any training data; they’re not used for anything else.

additionally, we’re adding the ability to flag an idea as “don’t send this to AI,” so you have another layer of security.

as a company, we’re SOC 2 compliant and pass security audits, all that good stuff.

security is 100% paramount. we’re talking with our largest enterprise customers and making sure we’re building this in a way their security departments are going to be thrilled with.

Dean Peters | 00:52:00–00:53:00
to your point, we’re seeing instances where organizations are saying “don’t use these online tools” like ChatGPT for a lot of the reasons Reg talked about.

when you start posting stuff into these tools, you’re not getting protections like pseudonymization and all that.

these are some of the risks that are in place.

I think it’s really timely.

now that we’ve got our problem space shaped, I think what we want to do is talk about shaping our solution space.

again, we have this template here: very similar brainstorming exercise, but to the right we have three additional guides for the conversation:

a problem statement,

a “how might we” statement,

and a solution statement.

Reg, go ahead and drive the bus.

Reg Cheramy | 00:53:00–00:56:00
here, we’re taking one of the problem statements from the previous exercise and diving deeper.

we click the AI button, and it automatically creates “how might we” questions, formatted intelligently, because we asked it to do that via the template.

it’s magically giving us several “how might we” scenarios, and then right behind that, some potential solutions.

Dean Peters | 00:56:00–00:57:00
yeah, I like that first “how might we”—it’s perfect.

this is dialed in: exactly an AI-powered pizza ordering platform.

what I like about this is you’re able to have a solution conversation that is not a bunch of wild animals trampling the product outcomes.

these recommended ideas are around the product outcome we’ve talked about, around the jobs-to-be-done we talked about.

when I’ve had to use GPT alone, I’ve had to continually remind the session because it runs out of tokens and forgets the original conversation.

Dean Peters | 00:57:00–01:00:00
Cameron, I think it’s time for our second poll now that we’ve seen how much work’s involved here, before we go to the lean canvas.

Cameron Lanier | 00:57:00–01:00:00
absolutely, let’s hop into that.

on a scale of 1 to 10, how effective has ChatGPT or similar tools been for you?

1 is “not effective at all,” 10 is “extremely effective.”

you might also think of it as: is the orange worth the squeeze?

someone notes in the chat “variation depending on application,” and that’s fair. try to average it out.

we’ll give it a few more seconds—I’ll start counting down: five, four, three, two, and one.

let’s share the results.

what are your thoughts here?

Reg Cheramy | 01:00:00–01:01:30
these are really intriguing results. it’s interesting to see a good chunk in that 6 or less range.

if we used the NPS style, that’s like the “this is frustrating” band.

ChatGPT can be frustrating because, unless you know how to talk to it effectively, it’s sometimes a struggle to get those answers.

at the top end, you see there are not many 10s, showing there’s some work to be done. but there is a good band in that 7–8 where people are finding it pretty effective.

I think if we crossed this data with the data of how long people have spent (from the first poll), my hypothesis is:

those who spent 50–200 hours are probably the ones at 7s and 8s,

those with 5–20 hours are probably in that 1–6 range.

it’d be interesting to actually cross that data.

Dean Peters | 01:01:30–01:03:00
this is great feedback.

I think Zoom is having issues with me today—it really doesn’t want me to do this webinar because we’re giving away secrets on how you can use generative AI to get real work done.

we’re going to talk a little about one of the things I teach: managing up and managing across.

we’ve done a good job of conversations with peers and teams about problem space and solution space, but now we’ve got to manage up and across:

managing up to our bosses looking at the financial picture,

managing across to other business units with different opinions.

we’re trying to get them to buy into this.

we’re going to use the lean canvas as a way of creating a consolidated conversation piece.

Reg, show how you use the guided lean canvas for this while I try to get Zoom working again.

Lean Canvas with Storm AI and Stakeholder Communication

Reg Cheramy | 01:03:00–01:07:30
awesome, thanks Dean.

lean canvas is one of the hundreds of templates we have built into Stormboard.

it has a neat feature called guides.

with guides, if you aren’t familiar with lean canvas, you can click the guide button and it teaches you how to fill in this problem statement.

it gives you questions, examples, and instructions for each section.

at the same time, we can use AI to do that.

if we go to the Storm AI prompt, you’re going to see this is very similar to the problem we just talked about, or the solution:

“a phone app that uses AI to recommend the most popular gluten-free pizza packages when I’m in a new town.”

that’s the context we’re giving it.

because lean canvas is a known template, Storm AI does a great job of this.

it quickly generates some of the problems you might have in that scenario, some solutions, key metrics you should think about, and so on.

Dean Peters | 01:07:30–01:10:00
for those who are technical product managers, pay special attention as it starts getting past items five, six, and seven—because that’s where those financial conversations show up that we might get flop sweat over.

we might not have MBAs, so AI can start those conversations for us as we manage up.

they want to know about total addressable and serviceable addressable markets: what am I looking at for cost, revenue, CAC, LTV—all that stuff we teach.

now you’ve been able to frame this conversation and of course modify it as well.

it reminds me of a podcast—Lenny’s Podcast—where he interviewed Melissa Perri (Escaping the Build Trap).

she joked that CPOs tell her they want to use the lean canvas. she says: the trick is not learning how to use the lean canvas—the trick is learning how to have the conversations to populate it, so it doesn’t become garbage-in, garbage-out.

I like what Stormboard is giving us here as a way of guiding us with examples.

Reg Cheramy | 01:10:00–01:13:30
one of the things I want to highlight is: the answers you’re getting, the section fills we’re doing with Storm AI, they’re shockingly good.

I’m going to tell you straight up: there’s no smoke and mirrors. I showed you the context we’re giving—it’s literally doing this with just that amount of information.

this is absolutely real-time.

the other thing: when we use this it’s good—these ideas are amazing—but when we bring our folks on and they add their ideas, then the AI is even better.

there’s another gear you’re not seeing here because most of these ideas are AI-generated; when you combine human and AI, it’s even more magical.

anyone who’s used ChatGPT would look at this and say, “wow, this is amazing; these ideas would take pages and pages of prompting,” and it’s just happening instantly in Storm AI.

I’ve actually taken this session and run it in GPT, and I can tell you it’s hundreds of lines, plus knowing markdown, etc.

we’ll share a link at the end of the session so you can see all the prompt engineering Dean had to do in pure ChatGPT.

something to think about: what we’re doing is something Dean has been teaching about AI for the past 10 years.

why don’t we let computers do what they do well—as a coaching system—so humans can do what we do well?

as a coaching network, Stormboard fits that bill: it’s that skybox coach helping you call the right play when your feet are on the ground.

Dean Peters | 01:13:30–01:16:30
speaking of the right play: I’m in a large enterprise and they want this in a Word document. they want almost a business case.

how might I take what I have and generate a communications document?

Reg, why don’t you show that?

Reg Cheramy | 01:16:30–01:20:00
here we’ve got a Microsoft Word document embedded directly inside Stormboard, generated automatically because it understands the context.

here are all the ideas by section—problems, solution, unfair advantage, etc.

then: all ideas by creation date, by creator, by legend color.

here are tasks assigned, incomplete tasks by due date and assignee, completed tasks, and top ideas by number of votes, plus a vote summary and comments.

we’ve turned visual collaboration into quantifiable, structured content.

with a two-second button push, you get a full stakeholder-ready Word document.

in any other tool, you’d be spending hours or days of manual transcription trying to turn screenshots into something usable.

Dean Peters | 01:20:00–01:21:00
I’ve had to do that. I know markdown, but how many of you know and master markdown to make this happen?

what Reg is doing here is what I love: letting the AI do the clerical work so humans can do the thinking.

Offers, Stormboard Giveaway, and Productside Announcements

Cameron Lanier | 01:21:00–01:24:00
I want to make sure we leave time for Q&A.

Reg, talk about how people can get Storm AI, and then we’ll share some Productside items.

Reg Cheramy | 01:24:00–01:27:00
one of the questions we had earlier was: how do we take advantage of Storm AI today?

we’re going to put in the chat a link to a form you can enter to win. only people on this call are eligible to win.

this is a $5,000 credit toward an enterprise subscription.

this gets you started—we’re giving away one of those and a bunch of $500 gift certificates to our enterprise subscription as consolation prizes as well.

that $5,000 value will include consulting with our experts, helping understand your problems, showing you how to solve them, and getting you some seats so you can explore it inside your organization.

if you go to the chat, Matt English just put that link in there. click on that and put your entry in, and we’ll get you into the draw.

if you’re still looking just to experiment with Stormboard, you can request a demo with our team and they’ll get on a call with you, even if you don’t win.

Cameron Lanier | 01:27:00–01:30:00
there are a couple things we want to discuss on the Productside.

first, we have a webinar coming up: “Why Do Products Fail?”

we’ve hinted at something we haven’t really talked about head-on: you saw today in Stormboard some of our fancy new templates.

behind the scenes, we’ve been working on new content, material, and instruction for product managers.

we were excited to marry that content with Storm AI to really take your product management workflow to the next level.

in our “Why Do Products Fail?” webinar, we’re going to release more information about those templates, including some of the ones you saw today.

next, you might recognize that beautiful face of Dean Peters—he’ll be teaching a workshop about the power of generative AI in product management.

again, it’s about marrying these tools: taking generative AI and using it with the framework we provide here at Productside.

Dean, you want to speak to that a bit more?

Dean Peters | 01:30:00–01:32:30
yeah, this is going to be a one-day workshop that helps you understand how to use generative AI in your work as product managers and how to include it in your products.

we’re going to touch on some of the topics we talked about today, as well as other topics.

you’ll see we have other courses coming up—digital product management, agile product management.

we’re going to help you take those solution-speak conversations and move them into the problem space.

Q&A: Jira Integration, Data Ownership, and Final Takeaways

Dean Peters | 01:32:30–01:36:00
I think now I want to talk with people and see what questions they have.

for example: “can you convert this information into stories?”

yes, we know we can—I’ve seen it happen in Stormboard and in ChatGPT.

so yes, you can create these into stories.

and yes, there will be a recording.

another question: “is there a way to edit specific prompts for the next section, like add additional context?”

we know you can do that in GPT via prompt engineering; in Stormboard, Reg showed a few ways of how you can add more context as you transition from section to section.

Reg Cheramy | 01:36:00–01:39:30
yeah, the beautiful thing is: we don’t want you to have to learn how to be a prompt engineer.

we’re doing a whole bunch of engineering on the back end.

to answer your question: you don’t actually have to edit prompts to make it more focused. you just have to add more ideas of the type you want, and it will automatically be more focused.

when you’re adding ideas to the different sections of Stormboard, it takes the ideas from all those sections to inform what it’s going to add for the next section.

it’s far more advanced than you doing the prompt engineering—it’s doing all of that automatically.

Dean Peters | 01:39:30–01:41:30
and again, if you’ve done prompt engineering in raw GPT, you know you have to maintain a running dialogue and keep feeding it back in.

there was another good question: “does it integrate with Jira?”

we talked about adding that into this mix but didn’t want to distract from the AI conversation.

Reg Cheramy | 01:41:30–01:44:00
yeah, we have an incredible bi-directional integration with Jira, Azure DevOps, and Rally.

when updates happen in Jira, they automatically move the ideas in Stormboard; when you move ideas in Stormboard, it automatically updates in Jira.

our Jira integration—if you’re interested—reach out to our team. we’ll give you a demo that’ll blow your mind.

we may talk with Productside about doing an upcoming webinar that dives deeper on that.

Dean Peters | 01:44:00–01:47:00
we’ve talked a lot about the strategic aspect today: if you don’t figure out the right thing to build, it doesn’t matter whether you build it right—you’re just delivering crap faster.

it would be great to have a more tactical conversation for the smoke jumpers on the front lines who have to get things done as well.

I see a question about data ownership—I know you talked about that a little bit, Reg.

Reg Cheramy | 01:47:00–01:49:30
yeah, all the data—you own it.

none of the data that goes into Storm AI or Stormboard is used to train OpenAI’s models or any other third-party generative models.

all of the data is 100% yours.

we’re doing everything we can to make the experience exactly what your security teams expect.

Dean Peters | 01:49:30–01:53:00
and as a typical enterprise SaaS product, you get the protections of multi-tenancy that you don’t get from open tools like raw ChatGPT or Bard.

plus, the collaborative nature: you get to collaborate with multiple people using AI, instead of one person typing in ChatGPT and then transcribing it.

for those of you worried that AI might replace your job: if all you’re doing is prompt engineering or slinging Jira tickets—doing feature-factory work—yes, AI could replace that.

what I like about Stormboard is it helps us frame conversations so we can be great negotiators managing up and across, great storytellers for teams and customers, and empower teams to be great decision-makers.

I’m really excited about using generative AI this way: taking out clerical work so we can be great conversationalists and spend more time in empathy interviews with our customers.

Reg Cheramy | 01:53:00–01:55:00
to answer another question: Storm AI is only available in the enterprise plan, not the small-business plan.

you can sign up for a 30-day free trial of our business subscription, and during that 30 days you’ll be able to access Storm AI.

after that, you’d need to convert to enterprise to continue to use it.

we’re happy to have a conversation and make sure we can bring value to your organization in whatever form.

Dean Peters | 01:55:00–01:57:30
for those of you who hung in here to the end, Reg is going to share the link to that ChatGPT session I created, so you can see all the prompt engineering I had to do—you have to double the number of lines you see there, because I had to do additional prep.

here’s a session I used to try running the same conversations, and you can see all the work I had to do.

it’s absolutely insane pages and pages of work I had to do traditionally, whereas with Stormboard you’re putting a two-liner in and Storm AI is doing all of it.

again, the question you need to ask yourself is: do I want to make a career as a prompt engineer, or do I want to make a career as a product manager having great conversations?

Cameron Lanier | 01:57:30–01:59:30
I appreciate that you guys shared that, because anyone who was in our previous webinar about ChatGPT saw the steps, saw how much it took, right Dean?

I love the fact this is so much faster. Dean loves prompt engineering; I do not. I just want to go there and get the answers.

the point today is: Storm AI is powerful; marry it with Productside’s templates and all of a sudden you’re getting to the answers you want more quickly.

and to Dean’s point about being dropped into a situation where you need to move fast—you have to have those two elements together: the template plus Storm AI.

Dean Peters | 01:59:30–02:01:00
absolutely. and we’re having requests for the previous webinar, so Cameron will share that link.

Reg Cheramy | 02:01:00–02:02:00
for previous independent consultants asking about possibilities, we’re happy to talk about how we can let you take advantage of this. just reach out to our team and we’ll figure out how to make something that really works for you.

Cameron Lanier | 02:02:00–02:04:00
any last questions before we get out of here?

Dean Peters | 02:04:00–02:05:00
all right folks: be awesome and help others do the same.

Cameron Lanier | 02:05:00–02:06:00
thanks everyone. I’m going to share the link real quick in the chat, and we will be sending out the link to this one as well.

thank you all for joining us, and have a great day.

Reg Cheramy | 02:05:00–02:06:00
awesome. thanks everybody.